Tuberculosis prevalence survey in Kashmir valley

Indian J Med Res 80, August 1984, pp 129-140 Tuberculosis prevalence survey in Kashmir valley S. Mayurnath, D.S. Anantharaman, G.V.J. Baily, M.P. Radhamani R.S. Vallishayee, P. Venkataraman* & S.P. Tripathy* Tuberculosis Prevention Trial & *Tuberculosis Research Centre, Madras Received November 11, 1983 A tuberculosis prevalence survey was conducted in about 18,000 persons in Kashmir valley situated about 1650 m above the mean sea level. All persons were tested with 3 IU of PPD-S and 10 units of PPD-B. Persons aged 5 yr and above were X-rayed (70 mm X-ray), and from such persons whose photofluorograms were interpreted as abnormal two specimens of sputum were collected and bacteriologically examined. In addition, a large X-ray of the chest was taken for children aged 0-4 yr who had reactions of 10 mm or more to PPD-S. They were then clinically examined by a paediatrician, taking into account all available data, for evaluation for any evidence of tuberculosis. The results of the survey showed that the prevalence of non-specific sensitivity (59%) in the Kashmir valley is significant. The prevalence of tuberculous infection was 38 per cent. The prevalence of culture positive tuberculous patients (3 per 1000) and that of abacillary X-ray positive patients (14 per 1000) were found to be similar in the two sexes contrary to the usual experience of a higher prevalence among males. Results from studies of phage typing, susceptibility to thiophen-2-carbonic acid hydrazide (TCH) and virulence in the guineapig of strains obtained from patients diagnosed in the survey showed that most of these strains belonged to phage type A, were resistant to TCH and were not of low virulence. Regional variations in the prevalence rates were seen, the- problem of tuberculosis appeared to he more in the Baramulla district as compared to Srinagar and Anantnag districts. A comparison of results obtained from the present survey with those obtained from the BCG trial in Chingleput (Tamil Nadu) revealed that the tuberculosis situation in the two areas was quite different. which had a high prevalence of non-specific sensitivity 1,2. A similar finding had earlier been reported by Bates et a1 3. However, in such areas of low prevalence of non- specific sensitivity very little is known about the pattern of tuberculous infection and disease. In order to obtain more In a survey of the distribution of nonspecific sensitivity in several parts of India it was observed that populations in villages situated at heights of 1200 m or more above mean sea level had a markedly lower prevalence of non-specific sensitivity than that in villages situated in the plains Present address: Indian Council of Medical Research, Ansari Nagar, New Delhi 110029 129

130 Tuberculosis prevalence survey in Kashmir valley information in this regard, a pilot survey for tuberculosis was conducted in the Kashmir valley, situated about 1650 m above mean sea level. The objective of the survey was to obtain estimates of the prevalence of tuberculous infection, nonspecific sensitivity and tuberculosis disease in all age groups. While designing the BCG trial it was considered necessary to obtain information on the protective effect of BCG vaccination in an area with low prevalence of non-specific sensitivity in addition to the study in Chingleput 4, where nonspecific sensitivity is highly prevalent. Thus, identification of a population with low non-specific sensitivity and information on tuberculous disease and infection in such a population would also help in planning such a study. Material & Methods The Kashmir valley consists of three districts : Srinagar, Baramulla and Anantnag. The survey was confined to the rural population (excluding boat population) and a simple random sample of villages was selected from each of the three districts in such a way that the size of the sample selected from each district was in proportion to the total population of the district. However, as it was thought desirable to include an urban area in the survey, Pantchok, a semi-urban area, on the outskirts of Srinagar town, was also included. In all, 7,679 persons from 11 villages in Srinagar district, 5,769 persons from 9 villages in Baramulla district and 5,307 persons from 9 villages in Anantnag district constituted the study population. Thus, the total population included for the survey was 18,755. Of these 18,311 constituted the de jure population and are considered for analysis. The locations of the 29 villages included for the survey are shown in Fig. 1. The survey was conducted from June to November 1978. The procedures adopted were briefly as follows. A complete census of each village was taken by a house to house visit and all persons registered. The left shoulder of each person was examined for the presence or absence of a BCG scar. Each person was administered two tests, one with 3 IU of PPD-S and the other with 10 units of PPD-B, on the mid-dorsal aspect of the two fore-arms allocated randomly. After 72 h, the transverse diameters of the indurations were measured, without knowledge of the order in which the two tests had been allocated to the two forearms. and recorded. At the time of testing, all persons aged 5 yr and above were offered an X-ray examination of the chest by 70 mm miniature photofluorograms. The X-ray films of the chest were read, independently, by two readers. All those persons showing X-ray abnormality in the lung or any part of the chest indicative of tuberculosis, pulmonary or otherwise, were eligible for sputum examination. From such persons, two samples of sputum-a supervised spot specimen and the other an overnight specimen were collected and subjected to bacteriological examination by fluorescent microscopy and culture. In addition, bacteriophage typing, thiophen-2-carbonic acid hydrazide (TCH) susceptibility 5 and virulence in the guineapig 6,7 of Mycobacterium tuberculosis strains isolated in the study were also carried out. All the bacteriological examinations were carried out at the Tuberculosis Research Centre, Madras.

Mayurnath et al 131 Fig. 1. Map showing locations of the selected villages. In addition, in order to obtain infor- including previous history of tuberculosis mation on the prevalence of tuberculosis and family history of tuberculosis. In in children aged under 5 yr, on the day of cases where further investigations were reading, a large X-ray of the chest was indicated (e.g., gland biopsy) these were taken for all children aged 0-4 yr who carried out. showed a reaction of 10 mm or more to PPD-S. These children were examined Results clinically by a Medical Officer and a Study population and coverages obtained : paediatrician for evidence of tuberculosis The population studied and coverages

132 Tuberculosis prevalence survey in Kashmir valley obtained for various examinations are shown in Table I, by age and sex. High coverages were obtained uniformly in all age groups and both sexes, with overall coverages ranging from 84-98 per cent for tuberculin testing, X-ray, sputum and clinical examinations. tion of persons with BCG scars was 22 per cent. Persons with a BCG scar were also included for the analysis except for the estimation of prevalence of infection and non-specific sensitivity. Prevalence of BCG scars : The area under Prevalence of infection : The distribution study is covered by the mass BCG vacci- of children aged 0-14 yr by size of reaction nation programme. BCG scars were seen to PPD-S is shown in Fig. 2. It can be in 4 (2613 examined), 17 (2492 examined), seen from the Fig. that the distribution is 37 (4645 examined), 29 (4169 examined) bimodal and the dividing line between the and 7 (2968 examined) per cent in the uninfected (left hand distribution) and age groups 0-4, 5-9, 10-24, 25-44 and 45 + the infected (right hand distribution), yr, respectively. For all ages, the propor- although not very sharp, is around 12 mm. Table I. Coverages obtained for various examinations by age and sex Sputum Clinical exam. Age group Sex Popn. regd. (yr) (de jure) test-read X-rayed elig. exam. elig. exam. 0-4 M 1410 F 1372 5-9 M 1303 F 1302 10-24 M 2621 F 2438 25-44 M 2477 F 2186 45+ M 1889 F 1313 1288 1268 1236 1216 2310 1994 1991 1657 1224 1233 10 1223 5 2390 19 2251 24 2137 67 2026 90 1726 187 1240 150 51 56 10 5 17 24 65 89 182 148 43 47 Total M 9700 8485 7486 283 274 51 F 8611 (87) 7908 (90) 6740 269 (97) 266 56 (92) (93) (99) Grand total 18311 16393 14226 552 107 (90) (92) Figures in parentheses show coverage (%) obtained; -Not eligible for examination

Mayurnath et al 133 The overall prevalence of infection was 38 per cent. The prevalence of infection rose rapidly, for both sexes, from the youngest age-group up to the age-group of 45-54 yr and then almost levelled off (Table II). At this age, over 85 per cent of males and over 75 per cent of females were infected. The prevalence of infection was about equal in both sexes up to the age-group 15-19 yr and thereafter lower among females than among males. Fig. 2. Distribution of 5892 children aged 0-14 yr by size of reaction to PPD-S. Prevalence of non-specific sensitivity : The distribution of children in three age-groups 0-4, 5-9 and 10-14 yr, by size of reaction to PPD-B is shown in Fig. 3. The distributions showed no clear separation between the uninfected and the infected even among children in the youngest age-groups. However, these distributions were skewed, the skewness starting around 10 mm. Therefore, for defining non-specific sensitivity 10 mm was taken as the limiting point. The overall prevalence was 59 per cent. The prevalence of non-specific sensitivity increased, in both the sexes, up to the age of 35 yr, by which age more than 95 per cent among males and more than 85 per cent among females were infected (Table III). The prevalence of non-specific sensitivity was lower among females after the age of 15 yr (P < 0.01). The slightly higher prevalence seen in females in the 0-14 yr age group was not statistically significant. Fig. 3. Distribution of children by size of reaction to PPD-B. Therefore, 12 mm was taken as a limit for defining reactors to PPD-S. Prevalence of disease in subjects aged 5 yr and above: Culture positive cases-the overall prevalence was 2.8 per 1000 2.6 per 1000 among males and 3.6 per 1000 among females (P>0.2). In both the

134 Tuberculosis prevalence survey in Kashmir valley Table II. Prevalence of infection ( > 12 mm to PPD-S) by age and sex Age group (yr) tested Males of reactors tested Females reactors 0-4 1223 25 2 1225 31 2 5-9 1008 66 7 1013 63 7 10-14 757 101 15 666 85 13 15-19 365 87 26 455 107 25 20-24 295 113 42 315 101 34 25-29 343 196 59 366 164 44 30-34 308 222 73 407 219 54 35-44 736 572 77 685 442 63 45-54 711 606 85 559 430 76 55-64 615 536 87 372 296 79 65+ 258 213 81 150 123 84 Total 6619 2737 42 6213 2061 33 *Proportion calculated after adjusting for the size of total (rural) population in each of the three districts, using the formulas : where P st =estimated proportion appropriate to stratified random sampling; P h =proportion of reactors (or cases) in the sample; N h =population in the district; N=total population in the three districts sexes, the prevalence rates generally increased with age, with a steeper rise after the age of 44 yr (Table IV). Abacillary radiologically possiblei/probable cases The prevalence of abacillary X-ray possible/probable cases (hereafter referred to as X-ray cases) has been estimated based on the readings of 70 mm photofluorograms. An X-ray case is defined as a person (abacillary) whose X-ray picture was interpreted as C or D (indicating possible or probable tuberculosis etiology respectively) by two readers. X-ray pictures interpreted as C or D by one of the readers only were submitted to a third (umpire) reader and persons for whom the X-ray pictures were then classified as C or D by the third reader are also considered as X-ray cases. The prevalence of abacillary X-ray cases is shown in Table V. The overall prevalence was 13.6 per 1000 13.3 per 1000 among males and 14.0 per 1000 among females (P>0.2). The

Mayurnath et al 135 Table III. Prevalence of non-specific sensitivity ( > 10 mm to PPD-B) by age and sex Age group (yr) tested Males of reactors tested Females of % reactors (adj)* 0-4 1223 5-9 1008 10-14 757 15-19 365 20-24 295 25-29 343 30-34 308 35-44 736 45-54 711 55-64 615 65+ 258 159 13 1225 175 15 268 29 1013 313 33 330 46 666 313 47 236 68 455 281 63 235 82 315 221 72 302 90 366 280 78 287 94 407 332 82 697 96 685 589 87 668 95 559 485 87 582 95 372 312 84 216 82 150 126 88 Total 6619 3980 61 6213 3427 56 * See foot-note under Table II Table IV. Prevalence of culture positive cases by age and sex Males Females X-rayed of cases Rate* per 1000 of X-rayed cases Rate* per 1000 5-14 2397 15-24 1226 25-34 1122 35-44 1015 45-54 800 55-64 658 65+ 268 0.5 2209 3 1.5 1.5 1265 2 2.1 2.0 1149 4 4.8 1.0 877 2 1.9 4.8 657 3 5.2 8.6 418 0 12.4 165 4 23.0 Total 7486 20 2.6 6740 18 3.0 *See foot-note under Table II

136 Tuberculosis prevalence survey in Kashmir valley Table V. Distribution of abacillary X-ray cases by age and sex Age group (yr) X-rayed Males Females of Rate* of Rate* cases per 1000 X-rayed cases per 1000 5-14 2394 1 0.2 2205 3 1.5 15-24 1221 3 2.8 1262 2 1.7 25-34 1117 7 6.0 1144 15 12.0 35-44 1010 14 10.5 872 17 18.3 45-54 794 13 17.6 652 16 25.1 55-64 648 30 47.3 415 24 63.6 65+ 263 29 124.8 157 14 100.2 Total 7447 97 13.3 6707 91 14.0 *See foot-note under Table II prevalence rates increased with age, in both the sexes, gradually up to the age of 54 yr, and more steeply thereafter. Differences in the prevalence of infection, non-specific sensitivity and disease in the three districts : As the prevalence of infection, non-specific sensitivity and disease rates, by age and sex, obtained in Pantchok were similar to those obtained in the rest of the material from Srinagar district the two sets of data were combined. Data on the prevalence of infection, non-specific sensitivity and disease is shown in Table VI, separately for the three districts. Minor variations in the prevalence rates for the three districts were observed. Thus, the overall prevalence of infection, for both sexes, was the highest in Baramulla district. This higher prevalence was confined to the age-group 5-34 yr among males and 5-24 yr among females. Considering the age-group 0-14 yr, the prevalence of non-specific sensitivity, for both males and females, was the highest in Anantnag district. The prevalence of culture positive cases was, in general, higher in the Baramulla district as compared to the other two districts. These rates are based on small numbers of cases. However, this difference was statistically significant (P< 0.01) among males but not among females. The prevalence of abacillary X-ray cases was similar in all the three districts and the small differences that were seen were not statistically significant. Drug sensitivity to isoniazid and streptomycin : Of the 38 patients with positive cultures, drug sensitivity results to both the drugs were available in 30 patients. Strains were regarded as resistant to isoniazid if the minimal inhibitory concentration (MIC) was more than 1 µg/ml and resistant to streptomycin if the resistance ratio (RR) was more than 4. Of the

Mayurnath et al 137 Table VI. Prevalence of infection, non-specific sensitivity and disease separately for the three districts Age group (yr) Srinagar Males Females exam. exam. Baramulla Males Females exam. exam. Anantnag Males Females exam. exam. Prevalence (%) of infection : 0-4 518 (2) 517 (3) 361 (2) 371 (4) 344 (1) 337 (1) 5-14 710 (6) 695 (7) 570 (14) 519 (12) 485 (9) 465 (8) 15-24 270 (23) 303 (20) 178 (40) 253 (38) 212 (32) 214 (23) 25-34 259 (59) 296 (51) 216 (71) 260 (52) 176 (64) 217 (46) 35-44 321 (79) 250 (68) 221 (77) 246 (67) 194 (77) 189 (57) 45+ 596 (87) 432 (79) 538 (84) 354 (78) 450 (86) 295 (79) Total 2674 (39) 2493 (31) 2084 (45) 2003 (37) 1861 (41) 1717 (31) Prevalence (%) of non-specific sensitivity : 0-4 518 (14) 517 (14) 361 (9) 371 (11) 344 (16) 337 (19) 5-14 710 (29) 695 (35) 570 (30) 519 (30) 485 (46) 465 (48) 15-24 270 (63) 303 (58) 178 (78) 253 (76) 212 (77) 214 (62) 25-34 259 (86) 296 (77) 216 (93) 260 (80) 176 (93) 217 (82) 35-44 321 (93) 250 (84) 221 (96) 246 (86) 194 (96) 189 (88) 45+ 596 (91) 432 (83) 538 (93) 354 (87) 450 (94) 295 (87) Total 2674 (56) 2493 (52) 2084 (60) 2003 (56) 1861 (65) 1717 (60) Prevalence (per 1000) of culture positive cases : 5+ 3116 (2.2) 2728 (1.5) 2323 (5.2) 2191 (4.1) 2047 (0.5) 1821 (2.7) Prevalence (per 1000) of abacillary X-ray cases : 5+ 3101 (12.3) 2716 (12.5) 2305 (13.4) 2179 (13.3) 2041 (13.7) 1812 (15.5) Rate is shown in parentheses 30 patients, 2 (7%) had strains resistant to TCH were undertaken for a total to both the drugs, 6 (20%) resistant to of 34 isoniazid sensitive strains, obtained isoniazid only, and the remaining 22 (73%) from 22 tuberculosis patients in the study had strains sensitive to both the drugs. population. Of these, 31 strains from 20 (91%) patients were classified as of phage Bacteriophage typing, susceptibility to type A. One strain (from one patient) TCH and virulence in the guineapig : Studies was of type B. The remaining patient, of bacteriophage typing and susceptibility who had given two strains, had a mixed

138 Tuberculosis prevalence survey in Kashmir valley pattern: one strain of type B and the other of type I. Of the 34 strains, 33 had an MIC of more than 20 µg/ml to TCH (i.e., resistant to TCH), while the other had an MIC of 10 µg/ml. Strains from 17 of the 22 patients were also tested for virulence in the guineapig. Only in three cases the mean root-index of virulence was less than 1 (0.80, 0.94 and 0.98). In all other cases it was more than 1, the overall mean being 1.24 (range 1.04-1.55). Prevalence of tuberculosis in children aged below 5 yr : Of the 2,556 children aged 0-4 yr and test-read for PPD-S, 107 children showed a reaction of 10 mm or more and were eligible for clinical examination. Of these, 90 were clinically examined by the medical officer and the paediatrician. A diagnosis of primary complex was made in 33 of the 90 children. No other forms of tuberculosis, glandular, bony, miliary or meningitis, were recorded. Thus, among children aged 0-4 yr, the prevalence of primary complex was estimated to be 15 per 1000. Chest X-rays of 29 of the 33 children diagnosed as having primary complex by the paediatrician were independently read by a tuberculosis specialist who had no knowledge of the observations of the radiologist or the paediatrician; 23 (79%) were interpreted as normal or non-tuberculosis by the tuberculosis specialist. The corresponding figure for the radiologist was 17 (59%). Thus, of the 29 children diagnosed as primary complex in as many as 16 (55%) the diagnosis was not confirmed by either the radiologist or the tuberculosis specialist. Even among the remaining 13 children where either of them interpreted the X-ray as of tuberculosis etiology (inactive or possibly active), in only 5 cases did they agree on the diagnosis. These results suggest that the estimated prevalence of primary complex in children aged below 5 yr may be an overestimate. Discussion A sample survey to obtain estimates of prevalence of tuberculous infection, nonspecific sensitivity and disease in Kashmir valley was carried out and the results obtained are reported. The prevalences increased with age. The prevalences of infection and non-specific sensitivity were more among males than among females. It is generally believed that the prevalence of non-specific sensitivity is very low in places at high altitudes. But the present survey showed that the prevalence of non-specific sensitivity in Kashmir valley is not insignificant. In the age-group 10-14 yr, nearly 50 per cent of the children showed a positive reaction to 10 units of PPD-B. Another significant finding of the survey was that the prevalence of culture positive cases and of abacillary X-ray positive cases among males were similar to those seen among females contrary to the usual experience that the prevalence of disease is 2 to 4 times higher in males than in females 4,8. Results from studies of phage typing, susceptibility to TCH and virulence in the guineapig of strains obtained from patients diagnosed in the survey showed that most of these strains belonged to phage type A, were resistant to TCH and were not of low virulence. Regional variation in the prevalence of infection, non-specific sensitivity and disease were observed in the three districts. Considering younger age-groups, prevalence of infection was maximum

Mayurnath et al 139 in Baramulla district and that of nonspecific sensitivity in Anantnag district. Considering the age-group 5+ yr, prevalence of culture positive cases was the highest in Baramulla district. These results suggested that the problem of tuberculosis was more in Baramulla district as compared to the other two districts. The survey in Kashmir valley was conducted by the same field teams conducting the Tuberculosis Prevention Trial 4 in the Chingleput area and using similar techniques. A comparison of the results from the two areas showed that the prevalence of infection and non-specific sensitivity, among both males and females, were considerably lower in the Kashmir valley. In the Chingleput area ho differences were observed in the prevalence of non-specific sensitivity between males and females whereas in Kashmir, after the age of 15 yr, females had a lower prevalence than males. The prevalence of culture positive cases was also higher in Chingleput area. But in Chingleput area the prevalence in males was nearly 4 times higher than that in females whereas in Kashmir the prevalences in the two sexes were similar. As a result, the higher prevalence in Chingleput area was mostly confined to males. Most of the strains from Kashmir area which were tested were shown to be of phage type A, resistant to TCH and with a mean rootindex of virulence of more than 1, whereas of 50 strains from Chingleput area (unpublished data), tested using similar procedures, 16 were of phage type A, 11 with an MIC of more than 20 µg/ml to TCH and 19 with a mean root-index of virulence of more than 1. Thus, the tuberculosis situation in the two areas appeared to be different. The differing ethnic characteristics, virulence of infecting organisms and immunological mechanisms as also environmental factors might have had a bearing on this difference. One of the objectives of the study was to know whether Kashmir valley could be considered to represent an area with a low prevalence of non-specific sensitivity and therefore suitable for conducting a study on the protective effect of BCG. The present survey has shown that although the prevalence of non-specific sensitivity in Kashmir valley is lower than that in Chingleput district it is still high and Kashmir valley may not be a suitable area for the proposed BCG study. We do not know whether populations located at still higher altitudes are free from non-specific sensitivity. Even if they are, the size of such populations would be too small for a BCG-trial, apart from other operational difficulties. Acknowledgment The authors gratefully acknowledge the unstinting work of the field teams of the Tuberculosis Prevention Trial and BCG teams of Jammu and Kashmir. The authors are particularly obliged to Dr Krishnalal Gupta DHS, Dr Tahir Mirza, STO, Jammu and Kashmir, for their guidance and support ; to Dr R.S. Minhas and his colleagues from the Department of Paediatrics, Srinagar; to Drs Pooran Raina, A.S. Sethi, Daljit Singh, Altaf Hussein, Sumangala Banu; Masood for their help in examining children; to Dr Hamid Khan, Director, TB Demonstration and Training Centre, Srinagar; to Drs P.S. Sikand and Sumeer Singh, Surgeons, Sher-I-Kashmiri Institute, Srinagar; to Dr S.N. Bhat, Radiologist, Women s Hospital, Srinagar for full cooperation. Authors also thank Dr S.P. Pamra, Director, New Delhi TB Centre, for reading large X-ray chest films of the children, and to Shri A. Venkatesha Reddy, for statistical assistance.

140 Tuberculosis prevalence survey in Kashmir valley 1. 2. 3. Bates, L.E., Busk, T. and Palmer, C.E. Research contributions of BCG vaccination programmes: Tuberculin sensitivity at different altitudes of residence. Public Health Rep (Wash) 66 (1951) 1427. 4. 5. References Raj Narain, Krishnamurthy, M.S. and Anantharaman, D.S. Prevalence of non-specific sensitivity in some parts of India. Indian J Med Res 63 (1975) 1098. Raj Narain, Vallishayee, R.S. and Venkatesha Reddy, A. Value of dual testing with PPD-S and PPD-B. Indian J Med Res 68 (1978) 204. Tuberculosis Prevention Trial, Madras. Trial of BCG vaccines in south India for tuberculosis prevention. Indian J Med Res 72 Suppl (1980) 1. Grange, J.M., Aber, V.R., Allen, B.W., Mitchison, D.A., Mikhail, J.R., McSwiggan, D.A. and Collins, C.H. Comparison of strains Mycobacterium tuberculosis from British, Ugandan and Asian immigrant patients: A study in bacteriophage typing, susceptibility to hydrogen peroxide and sensitivity to thiophen 2-carbonic acid hydrazide. Tubercle 58 (1977) 207. Mitchison, D.A., Wallace, J.G.. Bhatia, A.L., Selkon, J.B., Subbaiah, T.V. and Lancaster, M.C. A comparison of the virulence in guineapigs of south Indian and British tubercle bacilli. Tubercle 41 (1960) 1. Mitchison, D.A., Bhatia, A.L.. Radhakrishna, S., Selkon, J.B., Subbaiah, T.V. and Wallace, J.G. The virulence in the guineapig of tubercle bacilli isolated before treatment from south Indian patients with pulmonary tuberculosis. Bull WHO 25 (1961) 285. National Tuberculosis Institute, Bangalore. Tuberculosis in a rural population of south India: A five year epidemiological study. Bull WHO 51 (1974) 473. 9. Cochran, W.G. Sampling techniques (Asia Publishing House, Bombay) 1962 p 90 Reprint requests : Dr. S. Mayurnath, Deputy Director, Tuberculosis Prevention Trial Chetput, Madras 600031